Years of Graduate School
Daniel Margolis
Houston,TX
Summary
PhD candidate in Applied & Computational Mathematics (SMU, May 2026) with dissertation research integrating deep neural networks and stochastic volatility models for derivative pricing and high-frequency options trading. Published researcher (arXiv:2311.09456) with expertise spanning machine learning, optimization, signal processing, stochastic calculus, Monte Carlo simulation, and numerical PDEs. Production software experience from Lawrence Livermore National Laboratory. Extensive hands-on algorithmic trading strategy development across multiple platforms and asset classes. Proficient in Python (NumPy, Pandas, PyTorch, TensorFlow, scikit-learn), C++, MATLAB, and HPC environments. Seeking a Quantitative Researcher/Analyst role applying rigorous mathematical modeling, ML, and data-driven strategies to global financial markets.
Overview
6
6
2
2
Years of Financial Experience
3
3
Years of Machine Learning Experience
Work History
Research Intern
Lawrence Livermore National Laboratory
06.2023 - 08.2023
- Developed and documented 20+ computational examples for the SUNDIALS differential equation solver suite, a production-grade C/FORTRAN numerical library used across scientific computing and financial modeling applications
- Built testing frameworks and documentation pipelines (reStructuredText to HTML/LaTeX) for deployment across a 275,000-line codebase, applying production software engineering practices: version control, code review, and CI integration
- Modernized legacy FORTRAN codebases (F77/F90 to F2003) and collaborated with computational scientists under LLNL Fellow Dr. Daniel Reynolds on numerical methods for ODEs/DAEs
- Gained direct experience with numerical solvers applicable to modeling interest rate dynamics, credit risk, and derivative pricing frameworks
Teaching Assistant
Texas A&M University & Southern Methodist University
01.2020 - Current
- Instructed 12+ mathematics and computational courses including Numerical Methods, Differential Equations, Linear Algebra, Probability, and Calculus
- Mentored students in Python and MATLAB programming for computational methods; developed strong technical communication skills for conveying complex mathematical concepts to diverse audiences
Education
PhD - Applied & Computational Mathematics
Southern Methodist University
05.2026
MS - Mathematics
Texas A&M University
01.2022
BS & BA - Mathematics and Economics
Tulane University
01.2011
Skills
- Python
- NumPy
- SciPy
- Pandas
- Scikit-learn
- PyTorch
- TensorFlow
- XGBoost
- CVXPY
- C
- MATLAB
- FORTRAN
- SQL
- Git
- Linux
- API Development
- Machine Learning
- Neural Networks
- Gradient Boosting
- Bayesian Optimization
- Stochastic Calculus
- Monte Carlo Simulation
- Signal Processing
- Optimization
- Convex
- Portfolio
- PDE Solvers
- Finite Differences
- Spectral Methods
- Time Series Analysis
- Statistical Modeling
- Control Theory
- Derivative Pricing
- Black-Scholes
- Heston
- SABR
- Risk Analytics
- Greeks
- VaR
- Volatility Surface Construction
- Options Modeling
- Algorithmic Trading Strategy Development
- Pine Script
- EasyLanguage
- NinjaScript
- Backtesting
- Walk-Forward Validation
- Portfolio Optimization
- NVIDIA GPU SuperPOD
- HPC/Parallel Computing
- Data Pipeline Development
- Large-Scale Dataset Processing
- JSON/CSV Data Engineering
Technical Skills
Python (NumPy, SciPy, Pandas, scikit-learn, PyTorch, TensorFlow, XGBoost, CVXPY), C++, MATLAB, FORTRAN, SQL, Git, Linux, API Development, Machine Learning (Neural Networks, Gradient Boosting, Bayesian Optimization), Stochastic Calculus, Monte Carlo Simulation, Signal Processing, Optimization (Convex, Portfolio), PDE Solvers (Finite Differences, Spectral Methods), Time Series Analysis, Statistical Modeling, Control Theory, Derivative Pricing (Black-Scholes, Heston, SABR), Risk Analytics (Greeks, VaR), Volatility Surface Construction, Options Modeling, Algorithmic Trading Strategy Development (Pine Script, EasyLanguage, NinjaScript), Backtesting & Walk-Forward Validation, Portfolio Optimization, NVIDIA GPU SuperPOD, HPC/Parallel Computing, Data Pipeline Development, Large-Scale Dataset Processing, JSON/CSV Data Engineering
Research And Publications
- Dissertation: CDT-1D CNN Integration with Simpson-Sobolev Regularization for Options Trading, In Progress, Developing novel convolutional neural network architectures integrated with stochastic volatility models (Heston, SABR) for real-time derivative pricing and high-frequency trading signal generation across options markets., Implementing Simpson-Sobolev regularization to enforce smoothness constraints on neural network outputs, improving model stability and accuracy for pricing under volatile market conditions., Building end-to-end research pipeline: from problem formulation through data collection, model implementation (Python/PyTorch), backtesting, and risk-adjusted performance evaluation on large historical datasets.
- DeepMartNet: Martingale-Based Deep Neural Network for PDEs in ℝᵈ, Published, arXiv:2311.09456, Wei Cai, Andrew He, Daniel Margolis, Designed and executed large-scale neural network experiments on NVIDIA GPU supercomputer solving Fokker-Planck equations and Harmonic Oscillators in 5- to 200-dimensional spaces using PyTorch and TensorFlow., Developed scalable experiment framework leveraging martingale theory for eigenvalue problems - methodology directly applicable to derivative pricing and risk modeling via Feynman-Kac representations and signal extraction from high-dimensional financial data.
Algorithmic Trading Projects
- Developed and optimized quantitative trading strategies across TradingView (Pine Script), TradeStation (EasyLanguage), and NinjaTrader (NinjaScript) for futures and options markets.
- Implemented signal-driven approaches using geometric return analysis, volume categorization, momentum/volatility indicators, and compound return calculations with Bayesian parameter optimization.
- Achieved targeted profit factor improvements through rigorous backtesting, walk-forward validation, and risk-adjusted performance evaluation (Sharpe ratio, drawdown analysis) across multiple asset classes.
- Built custom indicators combining momentum, volatility, and volume signals for systematic trade entry/exit decisions.
Timeline
Research Intern
Lawrence Livermore National Laboratory
06.2023 - 08.2023
Teaching Assistant
Texas A&M University & Southern Methodist University
01.2020 - Current
PhD - Applied & Computational Mathematics
Southern Methodist University
MS - Mathematics
Texas A&M University
BS & BA - Mathematics and Economics
Tulane University